Sleep Apnea is not Associated with Elevations in Pulmonary Artery Pressure in a Clinical Cohort

Introduction Pulmonary hypertension (PH) is associated with adverse patient outcomes. Apneic events during sleep cause transient elevations in pulmonary artery pressure, however the association between sleep apnea and persistent daytime PH is less clear. Hypothesis We hypothesized that sleep apnea would be associated with elevated pulmonary artery systolic pressure (PASP) through both direct effects of hypoxia and indirect effects on left ventricular function. Methods We examined the electronic medical records of patients age ≥ 18 years undergoing sleep study within our large multisite hospital system between March 1999 and December 2015 who had an echocardiogram within 12 months of sleep study. Patients with severe left-sided valve disease (n=33), missing PASP (n=736), or missing apnea hypopnea index (AHI) (n=9) were excluded. An AHI ≥15 events/hour was used to define sleep apnea. PASP was noninvasively estimated by Doppler interrogation of the tricuspid regurgitant jet. Multivariable linear regression was used to examine associations between AHI and measures of left ventricular function and PASP. Results A total of 5,075 patients (age 58.2 ± 14.4 years; 53.0% female) were included. Sleep apnea was found in 45.4% of patients. Mean PASP was 33 ± 15 mm Hg with 34.8% having PASP ≥ 36 mmHg. Patients with higher AHI were older, more likely to be male, and had higher body mass index (BMI). Compared to patients without sleep apnea, those with sleep apnea had lower ejection fraction (EF) (53.6 vs 55.4%), higher left ventricular mass index (LVMI) (105 vs 92 g/m 2 ), and higher early mitral inflow to early diastolic mitral annual velocity ratio (E/e’) (11.9 vs 10.8). After adjustment for age, sex, site, and BMI, EF was 0.9 percentage points lower (95% CI [0.4, 1.4]), LVMI was 7.0 g/m 2 higher [5.1, 8.9], and E/e’ was 0.6 higher [0.1, 1.0] in patients with sleep apnea. In contrast, patients with and without sleep apnea had similar PASP (33.9 vs 33.2 mmHg; p=0.08). These findings persisted when limiting analysis to the 1,784 patients who underwent sleep study prior to echocardiogram. Conclusions In this analysis of real world patients referred for sleep study and echocardiogram, although sleep apnea was associated with markers of left ventricular dysfunction, we found no association between sleep apnea and PASP. This finding suggests sleep apnea does not directly elevate daytime pulmonary artery pressure. Elevations in PASP from left ventricular dysfunction may take time to develop and may not be apparent on cross-sectional analysis. Longitudinal follow-up of pulmonary artery pressure trajectories in patients with sleep apnea will provide greater insights on the impact of sleep apnea on PH.